Carlsen Supplies New Evidence That Chess Is A Draw With No Mistakes

Twinchicky   Most top players also think the Italian Game and the Two Knights are draws with best play.

Carlsen is actually supplying the opposite evidence.  The fact that Carlsen can "force" play his way against the best human players, but will lose hands down over and over to a strong engine, just proves that humans are nowhere near to proving anything about the ultimate solution to chess, even humans at the highest level...

Show me those games where he "lost hands down over and over to a strong engine". I'm not trying to disprove your words, I just want to see where it happened.

Well he hasn't, because there would be no point in even trying. There has been no expectation for a world chess champion to beat a computer engine in a formal tournament since Kasparov lost to the supercomputer Deeper Blue in 1997, SIXTEEN years ago. More recently in 2006 Fritz beat Kramnik easily whilst running on a laptop. And chess engines are much stronger now than they were in 2006. Their ratings are up around 3250 and Carlsen wouldn't stand a chance. And yet even these engines are still making suboptimal moves - an engine with a rating of 3500 would be able to beat them.

"Evidence" that observing draws means that we can prove that best play is always a draw is like saying that all dogs are brown, and then proving it by seeing lots of brown dogs. It fails even basic logical analysis.

arkledale

I'd guess it's a draw with perfect play.

But yeah, it's still just a guess. I don't see how any games or opening theory right now would prove anything. Carlsen is really good ofc, but in 20 years we might have computers with 5K+ playing strength, theory could be completely different from what it is now. We just can't say, maybe 5k elo computers will refute 1...e5 or the sicilian, who knows.

Sure there'd be a point. Since Carlsen is currently the highest rated person in history, seeing him head to head against Houdini would give us an example of the absolute limits of what the human mind can achieve.

And at the very least, we'd see some truly awesome playing. This match has shown that Carlsen needs to find a better opponent.

Isn't it obvious that chess without mistakes is just a draw? When you look at endgames with a computer, you see that as you go up in # of pieces on the board, the more variations there are that draw a game. In a King vs King and Pawn endgame 1 tempo is crucial so inaccuracies can be the same as a blunder, but if you make an inaccuracy in a board with more pieces like rook and 2 pawn vs rook and 3 pawn, you're usually ok. I hypothesize that there are probably dozens, maybe hundreds of "perfect" ways to play a chess game so that it ends in a draw. Of course, a game becomes stagnant without variance, so where is the variance? In our own imperfections.

To almost all grandmasters the game of chess is a draw. But this forum is not trying to give absolute proof that chess is a draw. Just additional evidence.

As the players become stronger there are more draws when they play against each other.  Look at the last 100 years of world championship matches and you will clearly see this trend. Again, this is not 100% proof--it is just evidence.

I don't read ponz111 as making a statement, rather asking a question which you write you are not strong enough to answer. Just because you can't answer his question doesn't 'change' his question into a statement; try to stay logical. He's asking a question which you can't answer. You're the one making a statement i.e.'Carlsen is making mistakes' which no matter how many times you repeat it doesn't make it 'somehow true'. Just tryna' help with some highschool logic here. 

Most people don't realize that IBM specially designed Deep Blue from the hardware up to calculate chess positions. This means that unlike normal computers, its chips were designed only for the types of calculations needed for chess. It was able to evaluate 200 million positions per second, which is still roughly 10 times better than the best desktops today!

Also, in these high profile matches, very few GMs have employed a true anti-computer style. But someone who was raised on computer analysis, such as Nakamura, knows exactly how to play the types of blockaded positions that make complete mockeries of all the engines.

MSC157 wrote:

Scotch? Or is it drawn too?

I wonder what weapon would Kramnik have next year when playing against Carlsen. 

.

They may as well hold Kramnik-Carlsen in Berlin, because that's all it will be, from both sides.

They can play right where the wall used to be.Same spot Roger Waters held his famous concert.

hehe, good!

"Evidence" can be circumstantial.  For example if you look at world Championship matches for the last 100 years you will see more and more draws.  Since the players are gradually getting better this is circumstantial evidence.

Also, it seems that Carlsen can hold to a draw the Ruy Lopez [one of the most respected and best openings] is an indication that the Ruy Lopez is a draw with best play which is in turn some circumstantial evidence that chess itself is a draw with best play.

Also the fact that just about every grandmaster in the world [with their hundreds of combined years of experience] believe chess is a draw with best play is circumstantial evidence.  It is not proof but it is evidence.

Also, the fact that a 100 billion games have been played and nobody is able to show one game out of 100 billion where White [or Black] won without his opponent making a mistake is very good circumstantial evidence that chess is a draw with best play.

Also, in Centaur Chess many believe the Petroff holds White to a draw if White opens 1. e4  e5  2. Nf3. 

 To me, the circumstantial evidence is overwhelming.  It is not 100% proof but still overwhelming.

Some circumstantial counter evidence. If we accept for a moment that engines do not make mistakes, then how is it that engine vs engine play regularly produces wins?

In practical terms, this notion of 'best play' is actually rather elusive. If best play is so complex that neither human intuition nor brute force calculation is able to find it, then arguing whether chess is drawn with best play is rather academic.

actually,even if carlsen's moves are suboptimal,he can still draw computers as he gets stronger.

arkledale,you never consider the human ability of getting stronger.

as a follow up, i would like to say that computers can't get stronger.

It is not certain at all that Carlsen makes mistakes every game.  By a mistake I mean a move which would change the outsome of the game unfavorably to him with best play by his opponent.

I saw only one such mistake he has made in the games with Anand. 

He is the top player in the world because he rarely makes a mistake.

"When you face opponents weaker than you" is the end of that sentence.

Sure you could add that but if you are about 100 points weaker, having White restores the balance. And you cannot make the same statement re other openings [for now at least]

Playing with the White pieces =/= a 100 point gain in ELO. If that were true, Black would lose a lot more in top-level play than it currently does.